CommsDay’s Friday exclusive on Google’s plans to form a trans-Pacific cable consortium gained much American media attention over the weekend—hardly surprising given that its stock hit an all-time record high the same day, valuing the company at an impressive US$174 billion.

This means Google accounts for over half of the total US$330 billion market capitalisation of US-listed internet information providers, out-shadowing the next largest in its sector—Yahoo, which is valued at just US$35 billion. Google is a true corporate phenomenon, and more valuable than the world’s most valuable telco, Verizon, by a differential of some US$50 billion.

One of the great things about having a story hit the American tech blogosphere is the instant feedback mechanism it provides—so as our article was cited by the likes of the New York Times, Techcrunch, Slashdot and Broadband Reports’ websites there was no shortage of reader and blogger comments on what it all meant.

Apart from the typical Google is “going-to-take-over-the-world-and-save-us-from-rapacious-telecoms/Microsoft” reactions, some took a more nuanced view—why on earth would Google be getting into the capacity commodity game by building its own cable?

Typical of this view was that expressed by Stuart Corner at ITWire where he wrote “If Google believes it can get a sustainable cost advantage by investing in a network, it may well do so. But there are a number of other new systems planned already. Technology advances rapidly. It seems more than likely that Google will be best able to meet its needs by relying on a healthy market with strong competition and investing its money in businesses that are centred around content and services and in which it can leverage its intellectual property and its massive access to end users, not raw connectivity.”

The problem with this view is that Google has, apparently, tried and failed persistently to get a satisfactory price on capacity from existing trans-Pacific cable providers. It understands the unit costs of fibre networks as it already owns such infrastructure in the continental United States and as the world’s Internet leviathan, is understandably frustrated that it can’t get a decent price on the trans-Pacific route—hardly surprising given that most of the capacity on such routes is controlled by the very Tier 1 telcos who see Google as a freeloader and unfair beneficiary of much of the value of the Internet economy.

Google doesn’t want to build a cable in order to sell bandwidth to third parties, although this  might be a consequence, but because simply it wants to control its destiny, especially as a voracious generator and recipient of Internet traffic.

And as our Friday report indicates, it doesn’t want to build the cable unilaterally, it wants to share the deployment costs with consortium partners—so it can get access to a fibre pair on a true cost basis, rather than paying marked-up retail rates.

Google is notoriously opaque in explaining its thinking to media and analysts. Indeed, when asked to comment on our Unity cable story it rather hilariously replied that there’s a lot of ocean out there! But it can’t be so opaque in its regulatory filings in support of its stock listing and it’s here we can see all the fears and paranoias of a company with runaway sales growth and a galactic stock price.

In the section of its quarterly filings dedicated to risks to the company, we can see the types of concerns that act as drivers for the cable plan.To quote, in part:

· “We rely on bandwidth providers, data centers or others in providing products and services to our users, and any failure or interruption in the services and products provided by these third parties could harm our ability to operate our business and damage our reputation.”

· “We rely on vendors, including data center and bandwidth providers. Any disruption in the network access or colocation services provided by these providers or any failure of these providers to handle current or higher volumes of use could significantly harm our business. Any financial or other difficulties our providers face may have negative effects on our business. We exercise little control over these vendors, which increases our vulnerability to problems with the services they provide.”

· “Our products and services depend on the ability of our users to access the internet, and certain of our products require significant bandwidth to work effectively. Currently, this access is provided by companies that have significant and increasing market power in the broadband and internet access marketplace, including incumbent telephone companies, cable companies and mobile communications companies. Some of these providers have stated that they may take measures that could degrade, disrupt or increase the cost of user access to certain of our products”

Need I go on? Google evidently sees putting as much of its network as possible under its control as an essential component to its continuing success.I would add something else. Google not only views the growing dominance of the major Tier 1 telcos over the Internet—the Verizons and the AT&Ts—as a threat to its future, but it also recognises that its continued growth depends on market conditions outside the US—already it banks an impressive 48% of revenues outside the US and this can only be expected to increase.

Increasing the total number of eyeballs and the amount of data they consume requires a healthy broadband access industry across the world. Google would tie that to the success or otherwise of the myriad of Tier 2 ISPs across the world in signing up customers and offering them high-octane services that keep up with broadband speed and data usage growth. Take any structural inefficiencies one observes in the US and multiply them by ten when it comes to the tough time faced by on-the-ground ISPs in Asia Pacific, especially in places where they never had a sustainable CLEC industry in the first place!

By building out its own networks, Google can directly peer with these Tier 2s and smaller foreign telcos, offering them favourable cost advantages on the current situation where they must pay expensive transit fees to connect to the Tier 1s in the US. This enables Google to act as a defacto Tier 1 in its own right, given the amount of traffic it would consequently control on-net. And it helps promote affordability of the Internet—from a supply side point of view—in countries across the Asia Pacific where Internet usage is still held back by the high costs of connecting to content and services housed in the United States.

Indeed, Google CEO Eric Schmidt told Wired Magazine in May that “one of the critical things for Google is for users to have inexpensive or free access to broadband.”

He also said in the interview that the advantage to owning one’s own network was that one can “tune it” and make it go faster, adding that the company’s very large data centres, will be the “small  ones” in a year or two.

It seems Google’s Unity plans are driven by the same incentives that drive its deployment of fibre and data centres across the United States.Except that an undersea fibre pair provides a lot of bandwidth—a terabit or so as a matter of fact—and this is where Google might consider becoming a secondary provider of connectivity.

Would it really say no to a Asia Tier 2 who might wish to peer with it to get onto the Google network and also asks, can we buy some transit from you instead of from those nasty US telcos? Why say no to the cash?

Another thing to note about Google is where it maintains its Asia Pacific operations. From what I can gather they are mainly concentrated in Japan, South Korea, China and Australia. This might give a clue as to where the proposed cable might head, or specifically in the case of Australia, where it might connect to given there might be as many three new cables already being built on the North-South route to Guam and Hawaii, all of which would gladly share the cost build if possible.

Our Friday report suggested that Google had been talking with Telstra and Asia Netcom—significant given that both plan major new cable installations.It’s less likely that Google would be discussing plans with its US “rival” Verizon, which is building another of the new trans-Pacific cables.And as we reported, the main player behind the Asia America Gateway cable—Telekom Malaysia—was apparently not invited to the recent Unity cable meeting in Sydney.

But it’s hard to make definitive comments about all who’s involved and who’s not. Only a dozen or so parties seem to be privy to the Unity plan and all would appear to have signed non-disclosure agreements. It is unclear if any party apart from Google itself would be privy to all its plans and all the potential partners it may be talking too.

So as they say in the classics, watch this space. But one thing is for sure. Google tends to leave lasting impressions on the markets it enters. It likes to set terms. Look at the astonishing concessions it gains from local authorities in the US for setting up shop with data centres, or its trading of terms for a guaranteed reserve bid at upcoming US spectrum auctions.

It would be naïve to think that Google might start taking more and more of its industry-leading content on-net without that having some form of impact on the existing economics of the Internet—not least because it may confer cost and strategic advantages that other large providers of content—the Yahoos and MSNs of this world—may seek to emulate.

by Grahame Lynch